LAUSR

Abstract English public speaking (EPS) is increasingly gaining prominence and popularity around the world, and this is especially true for university students in China. While self-efficacy is typically strongly correlated with language performance in general (Pajares & Graham, 1999), very little is known about self-efficacy and EPS performance specifically. Grounded in self-efficacy theory (Bandura, 1997) and research on English for academic purposes and English-as-a-foreign-language (EFL) literature, the main purpose of this study is to postulate and test against data a hypothesized model of EPS performance predictors to examine the relationships among student background characteristics, theoretically postulated sources of EPS self-efficacy, and the relationship between EPS self-efficacy and EPS performance. Participants were 82 EFL students enrolled in a university-level EPS course in China. Both quantitative and qualitative methods were employed for data collection and analysis. ANOVA results indicated substantial EPS self-efficacy and speech performance growth throughout the semester. Path analysis results provided evidence regarding hypothesized relationships among variables; qualitative data helped gain more fine-grained understanding of such relationships. These results add knowledge to self-efficacy theory in the EPS domain, provide a foundation for more robust models in other contexts, and affirm the importance of EPS instructional practices.